Architecting Constellations

Astel AI is a forward-thinking AI startup dedicated to building AI-first products for growing markets. Based on best-practice AI engineering and research, we create solutions for niche personal or corporate needs, with AI centered on both agentic and conventional deep learning approaches.

Explore Our Work →Get in touch

Our team previously built at startups backed by

01Work

Selected Case Studies

A closer look at how each constellation tackles a real problem, end to end.

Dendro · Consumer

Problem

Parents are drowning in scattered school emails: permission slips, event notices, and deadlines buried across dozens of unread threads, with no single place to track what needs attention before it's too late.

Approach

Nicety issues a private forwarding alias so parents never have to share direct inbox access. Nicey AI reads incoming school emails and extracts to-dos, events, and deadlines into an organized daily digest, calendar sync, and 24/7 chat assistant. It is encrypted end-to-end with AES-256 and built GDPR/CCPA compliant from day one.

Outcome

A live subscription product spanning Individual, Pro, Family, and School tiers, with WhatsApp digest delivery and support for multiple children across multiple schools.

AES-256 encryptionGDPR & CCPA compliant4 plan tiers

Visit Nicety →

02Scale

Engineered to Scale

We pioneer AI-first software for partners across many industries, architected on production-grade infrastructure designed to scale with every one of them.

We have experience building on Next.js and FastAPI for the application layer, Supabase for data, and LangChain and OpenRouter for orchestrating our models, deployed with Docker on Google Cloud behind Cloudflare and instrumented end-to-end with Jaeger and Opik.

On the machine learning side, we build and train models with PyTorch, track experiments with MLflow, run computer vision with Ultralytics, and handle classical ML with scikit-learn.

AI Workflows Built

AI Fields Covered

Computer VisionNatural Language ProcessingReinforcement LearningRoboticsMulti-Agent OrchestrationAgentic Workflows

03Team

The Team Behind Astel AI

Muhammad Ryanrahmadifa

Chief Technology Officer

Since 2022, Ryan has contributed to 4 VC-backed international startups (Y Combinator, Antler, East Ventures, Init-6, Temasek). His contributions are concentrated in the implementation of data science, ML/DL, computer vision, agentic AI, and multi-agent deep reinforcement learning. A graduate from Industrial Engineering at Institut Teknologi Bandung with research experience, he integrates systems thinking and science-based point of view to the world of AI Engineering.

4 years experience in Data Analytics
3 years experience in Machine Learning & Deep Learning
2 years experience in Agentic AI

Yasmin Salimah

Chief Product Officer

Currently enrolled into the Master's Program in Biomedical Sciences at Universitas Indonesia, Yasmin is an avid neuroscience enthusiast. She has contributed to research in neurodegenerative diseases using machine learning models in collaboration with Center for IoT Innovation (CITI) Laboratory, Taiwan Tech. She incorporates empathy and a holistic point of view in product development, and has been designing software as products since 2023. Her product design abilities have landed an award in healthcare software design, while continuously shaping her approach to user-centered innovation.

3 years experience in Product Design
1 year experience in Artificial Intelligence and Machine Learning

Our team previously built at startups backed by